1. Digitized by GoogIe
GENER AL CONSIDER ATIONS IN SIRE SELECTION
Sire selection is the key to any breeding program. All sire
selection procedures are designed to predict breeding value.
Performance of the individual himself, performance of ancestors
and collateral relatives, and performance of progeny are all
useful tools in sire evaluation. Their relative usefulness
varies d:!pending upon whether or not a trait can be measured or
estimated in the individual himself, on the heritability of the
traits of importance, and on prospective use of the sire or
sires selected.
If we set the figure 1.0 to represent complete accuracy in
predicting breeding value of a bull, a knowledge of relative
accuracies of prediction based on several types of information
are of use . For three levels of heritability some key figures
are:
1. From individual's own performance only:
Heritability
Accuracy of Breeding
Value Prediction
.20
.40
.60
.45
.63
.78
2. From progeny performance only:
Heritability
* Numbers of Progeny.
Accuracy of Breeding
Value Prediction
34
10* 20* 40* 80*
.20 .58 •72 .82 .90
.40 .73 .83 .91 .95
.60 .80 .88 .94 .97
2. Digitized by Google
3. From combined information on individual's own
performance and progeny performance:
Heritability
Accuracy of Breeding
Value Prediction
10* 20* 40* 80*
.20
.40
.60
.66
.80
.88
.75
.86
• 91
.84
• 92
.95
.90
.95
.97
3. Digitized by Google
*Numbers of progeny.
The accuracy of a breeding value determination applies only to
the potential sires evaluated under situations in which they
can be validly compared. Comparisons are of increasing
value when greater numbers are compared. Within-herd
comparisons are subject to less bias due to environmental
differences than are between-herd comparisons. The
usefulness of all sire evaluation programs isrooted in
widespread within-herd testing.
For most growth and conformation characters, heritability is
medium to high and the traits can be evaluated in the
individual. Whether or not later supplemented by progeny
tests, selection on the basis of own performance records
supplemented with records of ancestors and collateral
relatives should be intense. For most
traits, improved accuracy inestimating breeding value from
ancestor and collateral relative records is small . An exception
to this is weaning weight. Bulls from dams with consistently
good records are desired. For this trait a combination of
own record and an estimate of "Most Probable Producing
Ability" of the dam is recommended.
Lacking knowledge of heritability of between-herd
differences, within-herd differences should be the primary
criterion for selection on individual performance . Putting
bulls in central tests for the post-weaning period to
yearling age as a means of
minimizing effects of differences in herd environment is
desirable. Central testing facilities, preferably provided or
sponsored by breed associations, on a scale permitting testing of
the top five percent of the bulls would be desirable. The five
percent would
be selected based on within-herd performance to weaning.
Progeny testing is usually costly and can be justified only for
bulls of outstanding merit for traits measurable in the individual
35
4. Digitized by Google
himself. Emphasis in progeny testing should be on traits not
measurable intebulls themselves--carcasstraitsand maternal
ability of offspring.
Generally speaking,thecost of progeny testing canbejustified
only for selecting bulls to be used extensively in artificial
insemination or invery top seedstock herds.
Theforegoing relates largely to selection of bulls within herds
and points out the difficulties of making objective comparisons
of bulls raised in different herds. Generally speaking, the same
difficulties are encountered in comparing the progeny of bulls
each of which has progeny in only one herd.
A method for systematically producing offspring of some bulls
(termed reference sires) in many herds so that their progeny can
be directly compared with progeny of test bulls is outlined in the
material on a "National Sire Evaluation Program" which follows.
This isone method of developing sire comparisons with validity
across herds. Some breed associations now have such programs in
operation. They offer a means by which any breeder, large or
small, can identify germ plasm of potential usefulness inhis
herd. He can determine how his herd compares with others in the
breed through progeny testing one or more sires raised in or
being used in his own herd in comparison with reference sires,
36
5. Digitized by Google
NATIONAL SIRE EVALUATION PROGR AM
1
I. INTRODUCTION
A. Purpose and Scope
The purpose of a National Sire Evaluation Program
is to provide breeders with information on
"Expected Progeny Differences"between bulls.
"Expected Progeny Difference" is the best
estimate possible from available data of the
difference between the average of a large sample
of abull'sprogeny fromrepresentative cows as
compared to progeny of base reference sires when
bred to similar cows. The expectation is that
information on "Expected Progeny Differences"
will aid breeders in making decisions on
selection of bulls best suited to accomplishment
of specific objectives for the herd. A
secondary purpose is to enable breed associations
or other sponsoring organizations to measure the
direction and magnitude of genetic changes in a
breed over time.
Focus of the program should be on measurable
j characters related to the economic production of
quality beef.
A National Sire Evaluation Program for any breed
should be planned and conducted by an organization
not having direct interests in any specific animal
under test. Breed associations may sponsor
programs or they maybe sponsored by private or
public organizations with interests in more than
one breed. It is in the interests of all
concerned that there not be more than one program
per breed. Regardless of whether the sponsoring
organization is conducting programs for only one
breed or for several, each program should be
nationwide with "Expected Progeny Differences" and
related information to be on a within-breed basis.
gReport of National Sire Evaluation Committee adopted by
Beef Improvement Federation, April 1971.
37
6. Digitized by Google
B. Summary of Program
Beef Improvement Federation guidelines for a
National Sire Evaluation Program include as a
first step the encouragement of herd performance
testing as a means of identifying bulls with
desired performance characters: Records of
individuals ranking high within herds in 205-day
weaning weight and 365-day weight will be
published for use by other breeders primarily as
an aid in making decisions relative to within-
herd selections for progeny testing, use in
purebred herds or for commercial use.
Possibilities for meaningful between-herd
comparisons will be very limited in the early
stages of a program. Later, as ties are
established between herd sires and the reference
sires used in progeny testing programs, between-
herd comparisons of greater validity will become
possible.
Two procedures for progeny testing are outlined.
Thefirst isforwithin herd use. It does not
provide for comparisons with sires in other
herds. The other involves use of designated
reference sires in either single-herd or
multiple-herd tests. This procedure permits
breed-wide comparisons of bulls under progeny
test.
Emphasis in these guidelines is on principles
which will permit individual breeds to adapt the
program to their specific needs. Traits for
which procedures are outlined include 205-day
weaning weight, 365-day or 550-day yearling
weight, carcass weight per day of age, carcass
yield of preferred retail cuts expressed both
as a percentage of carcass weight and per day
of age, carcass quality grade, cm1 maternal
qualities and progeny testing for deleterious
recessive genes. Progeny testing can be
sequential with individual breeders and/or
sponsoring groups to select the traits to be
evaluated in specific programs. Programs need
not be limited to traits discussed in these
guidelines. The program calls for publication
of results and calculation of "Expected Progeny
38
7. Digitized by Google
Differences" for 365-day weight, USDA carcass
quality grade and carcass yield of preferred
retail cuts per day of age.
II . INDIVIDUAL BULL PERFORMA..CE EVALUATION
In a National Sire Evaluation Program widespread programs
of within-herd performance testing in the purebred herds
of a breed are a prerequisite. These records identify
high ranking individuals within herds, i.e. potential
candidates for progeny testing or for immediate use in
seedstock herds. In addition to individual performance
records, all available information on sire, dam, and sibs
should be utilized to estimate "Expected Progeny
Differences" with maximum accuracy possible from the data.
Initially, between-herd comparisons uill be of limited
value due to lack of knowledge of genetic differences
between herds. Also, therewill be fewdirect tieswith
other herds .
As the pr_ogram progresses, the progeny test program
(involving reference sires)will develop information on
genetic differences between herds and will also involve
direct and indirect ties with other herds. These things,
together with within-herd performance records will
increase validity of between-herd comparisons.
j
Procedures with some background material for evaluating
and publishing individual evaluations are:
A. Weaning weight
Weaning weight is included as part of the report
on a bull as an aid in evaluation (1)since it
is a part of yearling weight, and (2)as an
early indication of the possible maternal
performance of his daughters. Weaning weight
will be evaluated by BIF procedures and
expressed as 205-day weight. Emphasis for
weaning weight will be on ratio of the
individual bull's 205-day weight to the average
of his contemporaries in the same herd.
B. Yearling weight and carcass yield
Yearling weight combines in a meaningful way the
growth of an animal over at least two distinct
39
8. Digitized by Google
management regimes. It should be evaluated
and expressed by BIF procedures as either 365-,
452-, or 550-day weight. Post-weaning tests
may be conducted according to BIF procedures
either in herd of origin or in a central bull
test. Breeders with fewer than 10 contemporary
bull calves in their own herds should arrange
to test collectively with other breeders in
order to participate in a National Sire
Evaluation Program.
Methods for estimating carcass yield of live
animals are not considered sufficiently
accurate nor consistent from location to
location to justify their inclusion in
individual evaluations at this time. However,
when and if technology permits.live animal
evaluation of potential carcass yield should
be incorporated for each bull at the
conclusion of the post-weaning test.
Publication of individual performance records
isoptional. If the breeder elects to publish,
material to be published will include:
1. Identification
Breeder, owner, sire, dam, birth date, age
of dam, State which raised, State in which
post-weaning test conducted,and whether
post-weaning test was a single-herd or
central test.
2. 205-day weight information
Adjusted 205-day weight.
Ratio of adjusted 205-day weight to average
of contemporaries.
Number and averages of contemporaries .
3. 365-, 452-, or 550-day weight information
atio of adjusted 365-, 452-, or 550-day
weight to averages of contemporaries
from same herd. (If post-weaning test
ina central test,ratios as above to
average of all animals in test.)
40
9. Digitized by Google
Number and average of contemporaries
from same herd. (If tested in central
test, number and average of all
animals intest.)
When programs have advanced to the point that
"Expected Progeny Differences" based on progeny
are available for sires of performance tested
bulls, then "Expected Progeny Differences"
shall be calculated for them and presented with
prediction errors for 365-day weight, USDA
carcass quality grade, and carcass yield of
preferred retail cuts per day of age.
III. PROGENY TESTING FOR GROWTH AND CARCASS CHARACTERS
Generally speaking, progeny testing cannot be justified if
it is solely for the purpose of choosing among bulls
evaluated for growth in the same herd. Hoever, progeny
testing is the only accurate means now available for
comparing bulls which are not contemporaries. It i5 the
only method for evaluating carcasses.
Progeny tests can be designed to provide any desired level
of prediction error (Appendix 1). Numbers of females in
test herds are usually a limiting factor. Thus, decisions
which will optimize use of test herds must be made between
numbers of bulls to be tested and prediction error of
individuals tested.
A. General Rules for Progeny Tests
1. All progeny tests shall be planned in
advance and plans approved by the
sponsoring organization.
2. The sponsoring organization must develop
appropriate procedures for determining
that cows within group (group defined as
cows of a given breed or cross managed as
a single herd or unit) are randomly
allotted within age to the bulls under
test, that cows are bred as planned, that
birth dates are promptly and accurately
recorded, that progeny are managed either
uniformly or in a stratified fashion so
41
)
10. Digitized by Google
that all sire groups are represented in
each management situation or adequate
ties provided, and that records are
taken as prescribed.
3. Meaningful progeny tests can be
conducted only when two - bulls
are tested.
4. Deviations from any of the items listed
in 2 (above) are serious and result in
biased sire comparisons.
Two types of progeny test are possible, both are
useful, and both should be part of a National
Sire Evaluation Program. The first is termed a
!!Breeder Test" in which there are no ties to
other herds or groups and progeny comparisons
can be made only within the test. The second is
termed a "Reference Sire Test11
in which ties to
other tests make comparisons on a national basis
possible.
B. Breeders Tests
Breeders may test as few (two minimum) or as
many sires as they wish for the traits they
designate. Bulls in this type of test are ranked
by contemporary comparison. Bulls with progeny
in different tests and with no ties to other
tests cannot be compared. Each breeder is
allowed to choose the number of progeny from
each bull (hence, to determine the prediction
error of the comparisons)and may have many
progeny from some bulls and few from others.
The sponsoring organization will summarize and
analyze results of these tests and return to
breeder. Advantages of this test are that it
may be entirely by natural service if desired
and that if reference sire progeny are not
wanted in a herd, none need be produced.
The principle disadvantage of the breeder test is
that comparisons can be made only among the bulls
tested. No comparisons with bulls in other herds
are possible. If the test is conducted in only
42
11. Digitized by Google
one herd (as would usually be the case), the
degree to which results apply generally will
not be known. Because bulls used in some
herds will be of substantially higher merit
than bulls used in other herds, the sire
values from breeder's tests cannot be used
directly to rank bulls from different herds
without bias. Direct use of these sire values
would favor bulls compared in the same herd
with poor bulls and discredit good bulls used
in the same herd with other good bulls.
Sires will be evaluated by appropriate least
squares procedures.
c. Reference Sire Test
The obvious solution to the principle problem
of the "Breeder Test," namely, that comparisons
cannot be made between tests, isto include in
each breeder's test one or more reference bulls;
bulls who are also used in other herds and can
link together the various breeders'bulls. The
criterion for ranking breeders 'bulls is the
"Expected Progeny Difference" beten breeders '
bulls and the base reference sires. This
provides an unbiased ranking of breeders'bulls
(see Appendix 2). A national ranking requires
that all sires be compared directly or
indirectly with one or more sires designated by
the sponsoring organizations as base reference
sires. The criterion for ranking breeders'
bulls on a national basis is:
(Breeder's bull - reference sires) +
(in breeder's herd)
(Reference sires -
all base reference
siresused in
breeder's herd)
(all other herds)
Prediction error is measured as the square root
of the sum of the expected sampling variance of
the comparison (see Appendix 1). If the
reference sires have many contemporary progeny,
this prediction error should not be appreciably
more than a breeder test prediction error.
43
,j
12. Digitized by Google
This procedure for a national ranking of
progeny tested bulls recognizes that
unknown genetic and management differences
between herds are large, yet allows unbiased
ranking through carefully designed
comparisons in either single-herd or
multiple-herd tests. In a single-herd test,
the disruption of breeder's management
program is minimal as he can continue to
breed most of his cows naturally if he
desires, requiring only that a representative
group of cows in each herd be mated
artificially to reference sires.
Multiple-herd testing is to be preferred. It
requires that bulls under test produce
progeny ina number of herds in which
reference sires also produce progeny.
Multiple herd testing provides information of
more general applicability if genetic-
environmental interactions should be
important. Further, multiple-herd testing
reduces chances for biases of a non-random
nature to influence results.
Results will be analyzed and sunnnarized by
appropriate least square procedures (see
Appendix 3).
Prediction errors of "Expected Progeny
Differences "will depend upon numbers of
progeny per tested sire and numbers from
reference sires in the herd(s) when direct
comparisons are made.
As more bulls are tested in either a single-
herd or multiple-herd test, it is important
to increase the number of progeny from
reference sires. Tentative numbers to be
required are:
44